MCP Project Setup

MCP Project Setup is an MCP server starter template that provides boilerplate configurations and sample implementations for building Model Context Protocol services with Python and FastMCP. The repository connects to external APIs like OpenWeatherMap and provides ready-to-run examples, including a greeting tool, resource handlers, and a live weather query tool. Developers and AI engineers use this project to learn MCP architecture, test server components using the MCP inspector, and connect custom local tools directly to LLM command-line interfaces like the Gemini CLI. By providing predefined Python scripts for tools and resources alongside environment configuration files, the project eliminates repetitive boilerplate setup. Developers can run development servers locally, test function endpoints via command-line arguments, and integrate standard stdio-based MCP servers into external AI clients by configuring JSON settings with executable paths and necessary environment variables.

Category: Other & General Purpose

Tags: boilerplate, starter, template, weather

Visit MCP Project Setup

How to install and configure MCP Project Setup

1. Create a Conda environment and install dependencies: bash conda create -n mcp python=3.12 -y curl -LsSf https://astral.sh/uv/install.sh | sh conda run -n mcp uv pip install fastmcp 2. Configure your OpenWeatherMap API key in a .env file: env OPENWEATHER_API_KEY=YOUR_API_KEY_HERE 3. To integrate with Gemini CLI, locate ~/.gemini/settings.json (or %APPDATA%\gemini\settings.json on Windows) and add the server configuration under mcpServers: json { "mcpServers": { "weather_server": { "command": "uv", "args": ["run", "/absolute/path/to/mcp/mcp_weather.py"], "cwd": "/absolute/path/to/mcp", "timeout": 10000 } } }

What you can do with MCP Project Setup

• Testing custom MCP tools and endpoints locally using the FastMCP developer inspector interface before client deployment. * Fetching real-time weather information from OpenWeatherMap through an integrated MCP weather tool inside Gemini CLI sessions. * Prototyping basic Model Context Protocol greeting interactions using the included Hello World starter script. * Serving and testing custom structured data resources using standard MCP resource definitions and testing flags. * Bootstrapping new Python-based MCP projects using an isolated Conda and uv environment setup.

Key facts

• https://github.com/MarkAndersonIX/mcp-gemini-cli-base
• Other & General Purpose
• boilerplate, starter, template, weather

Part of MCP Servers

How do I install MCP Project Setup?

You install the project by cloning the repository, creating a Python 3.12 Conda environment, and installing the fastmcp package using uv pip install fastmcp. You also need to configure any necessary environment variables like OPENWEATHER_API_KEY in a local .env file.

What can MCP Project Setup do?

It serves as a boilerplate project providing working examples of FastMCP servers. It includes a basic Hello World tool, an MCP resource definition script, and an OpenWeatherMap server that delivers real-time weather data to compatible language model clients.

Which MCP clients work with MCP Project Setup?

The project includes explicit configuration instructions for integrating with Google Gemini CLI via its settings.json configuration file. Because it exposes standard MCP servers through FastMCP, it can also connect to other stdio-compatible MCP clients and the FastMCP developer inspector.

How do I test tools in MCP Project Setup?

You can inspect and test tools visually by launching the FastMCP inspector with fastmcp dev mcp_hello.py:mcp, or run automated script tests directly from the command line using uv run mcp_weather.py --test.